Human dendritic cells will be-used to generate effective T cell-mediated immunity against melanoma. Dendritic cells are critical to the onset of cell-mediated immunity, and they can be generated in substantial numbers from CD34+ progenitors for large scale experimental evaluation and therapeutic applications in vivo. Melanoma expresses well-defined antigens to which T cell responses occur, and immunotherapeutic interventions have yielded objective responses. Clearly, if manipulation of T cell anti-tumoral immunity is feasible in humans, dendritic cells are the best costimulatory antigen-presenting cells to use, and melanomas offer the best defined tumor antigens to elicit antigen-specific, MHC- restricted T cell responses. Dendritic cells will be expanded from CD34+ bone marrow progenitors under the aegis of c-kit-ligand, GM-CSF, and TNFalpha. Dendritic cell capture and processing of defined MAGE- and tyrosinase.derived polypeptides and peptides will be evaluated during dendritic cell expansion from immature progenitors. Effective presentation will be monitored by HLA-restricted responses of antigen- specific clonal cytolytic T lymphocytes. Processing and presentation of these melanoma antigens by differentiating dendritic cells will be compared to their concomitant levels of costimulatory ligand expression and accessory cell function. Primary T cells from tumor-bearing hosts will be stimulated with melanoma antigens presented on dendritic cells that simultaneously deliver important costimulatory signals. Effective stimulation will be confirmed by cytolytic activity directed against targets bearing the eliciting melanoma Ag and the appropriate class l MHC restricting determinant. Cytokine enhancement of dendritic cell function or T cell reactivity will be evaluated. Blocking studies will be performed to identify the critical costimulatory signals that optimize T cell responsiveness to melanoma antigens. The relevance of these in vitro responses to tumor protection will be determined using melanoma Ag- bearing dendritic cells as in vivo immunogens to generate protective tumor-specific CTL in melanoma patients. The experiments outlined constitute an ideal program in which the biologic mechanisms underlying cell-mediated immunity to human cancer can be successfully defined for potential therapeutic interventions.